A handle to schedulers paused on a barrier.

Public trait implemented by every packet batch type. This trait should be used as a constraint for any functions or places where a Batch type is requi

An interface implemented by all batches for iterating through the set of packets in a batch. This is private to the framework and not exposed. # Safe

`CompositionBatch` allows multiple NFs to be combined. A composition batch resets the packet pointer so that each NF can treat packets as originating

This scheduler is designed to allow NetBricks to be embedded in other vswitches (e.g., Bess). As a result it neither does any of the resource accounti

Null metadata associated with packets initially.

A trait implemented by all headers, used for reading them from a mbuf.

Results from inserting into `ReorderedBuffer`

A multiproducer single consumer queue for mbufs. The main difference when compared to `std::sync::mpsc` is that this does not use a linked list (to av

`NetBricks` control configuration. In theory all applications create one of these, either through the use of `read_configuration` or manually using ar

Base packet batch structure, this represents an array of mbufs and is the primary interface for sending and receiving packets from DPDK, allocations,

Generic trait for objects that can receive packets.

Generic trait for objects that can send packets.

A struct containing the parsed information returned by the `PayloadEnumerator`.

An enumerator over both the header and the payload. The payload is represented as an appropriately sized slice of bytes. The expectation is therefore

A DPDK based PMD port. Send and receive should not be called directly on this structure but on the port queue structure instead.

Configuration for each port (network device) in `NetBricks`.

Statistics for PMD port.

A structure for storing data that might be received out of order. This allows data to be inserted in any order, but then allows ordered read from data

A ring buffer which can be used to insert and read ordered data.

Used to keep stats about each pipeline and eventually grant tokens, etc.

Used to keep stats about each pipeline and eventually grant tokens, etc.

Messages that can be sent on the scheduler channel to add or remove tasks.

A linked list of segments, this tries to avoid allocations whenever possible. In steady state (regardless of bad choices made by the developer) there

FIXME: While WIP

A very simple round-robin scheduler. This should really be more of a DRR scheduler.